Erschienen in:
11.03.2019 | Cardiac
Epicardial fat volume measured on nongated chest CT is a predictor of coronary artery disease
verfasst von:
Yasunori Nagayama, Naoki Nakamura, Ryo Itatani, Seitaro Oda, Shinichiro Kusunoki, Hideo Takahashi, Takeshi Nakaura, Daisuke Utsunomiya, Yasuyuki Yamashita
Erschienen in:
European Radiology
|
Ausgabe 7/2019
Einloggen, um Zugang zu erhalten
Abstract
Objective
To investigate whether epicardial fat volume (EFV) quantified on ECG-nongated noncontrast CT (nongated-NCCT) could be used as a reliable and reproducible predictor for coronary artery disease (CAD).
Methods
One hundred seventeen subjects (65 men, mean age 66.6 ± 11.9 years) underwent coronary CT angiography (CCTA) and nongated-NCCT during a single session because of symptoms suggestive of CAD. Two observers independently quantified EFV on both images. Correlation between CCTA-EFV and nongated-NCCT-EFV was assessed using Pearson’s correlation coefficient and Bland–Altman plots. Inter-observer agreement was analyzed using concordance correlation coefficients (CCC). Coronary risk factors including EFV were compared between CAD-positive (> 50% stenosis) and CAD-negative groups. The association between EFV and CAD was analyzed using multivariate logistic regression. ROC analysis was performed, and AUC was compared with DeLong’s method.
Results
Seventy-four subjects were diagnosed with CAD. An excellent correlation was noted between CCTA-EFV and nongated-NCCT-EFV (r = 0.948, p < 0.001), despite the systematic difference between both measurements (mean bias, 1.26). Inter-observer agreement was nearly perfect (CCC, 0.988 and 0.985 for CCTA and nongated-NCCT, respectively, p < 0.001). Significant differences were noted between subjects with versus without CAD in age, hypertension, and EFV on both types of images (p ≤ 0.026). Multivariate analysis revealed that increased EFV on CCTA (odds ratio 1.185, p = 0.003) and nongated-NCCT (odds ratio 1.20, p = 0.015) was independently associated with CAD. There was no significant difference between CCTA-EFV and nongated-NCCT-EFV in AUC for the prediction of CAD (0.659 vs 0.665, p = 0.706).
Conclusions
Despite the absence of ECG gating, EFV measured on NCCT may serve as a reproducible predictor for CAD with accuracy equivalent to EFV measured on CCTA.
Key Points
• Despite the absence of ECG gating, the EFV on NCCT provides nearly perfect inter-observer reproducibility and shows excellent correlation with measurements on gated CCTA.
• EFV on nongated-NCCT may serve as an independent biomarker for predicting coronary artery disease with accuracy equivalent to that of EFV on gated CCTA.